ALLEN GLACIER AND OTHER ICE TONGUES 445 The front of Allen Glacier has a distinct crenulation which is interpreted not as due to lobation of the ice front, but to slumping of areas where the ice melts out. One of these has already been spoken of as containing a good-sized pond. Another valley SO or 40 feet deep was formed between our visits in 1909 and 1910 and in the latter year its margins showed long cliffs of bare black ice and a floor 100 by 800 yards strewn with dead alder. Earlier events of this sort were indicated by deep valley-like crenulations in the glacier front, some of them with mature alder growing on the sides where slumping had ceased, others with the slopes still covered with dead or dying vegetation. Relation of Allen Glacier to the Railway. Nowhere else in the world, so far as we know, is a railway built for 5$ miles upon the end of a living glacier (Pis. CLXXI, CLXXXI, B). The ballast beneath the ties and rails of the railway actually rests upon the ice, not upon an abandoned moraine as at Heney Glacier. Since the railway was built, disadvantages in this location have been revealed, notably the settling of the track as the ice beneath it melts, the frequent breaking out of new streams, calling for new trestles, and in one case the shifting of a bridge support 18 inches toward the river, necessitating the driving of a new pile. These difficulties will recur every summer and will always render this section of the railway expensive to maintain and slow for trains to run over. Of course there is no danger to passengers on a railway thus located, provided track-walkers watch the railway grade carefully. The gravest problem in connection with the section of the railway on Allen Glacier comes in connection with a possible advance of the glacier. This would destroy the track and completely block traffic on the whole railway, for there is no way to go around. After a period of immunity of 67 years such an advance is likely to come at any time and may be especially imminent any time within the next few years,1 for the adjacent Childs, Grinnell, and Heney Glaciers had advances in 1910 and 1911, suggesting that the snow-fields west of Copper River are likely to also produce an advance in the Allen Glacier, which lies between the Grinnell and Heney ice streams. Such advance, however, would probably be short-lived and, if not too large, its impetus might all be taken up in the interior flat, so that the outer edge of the glacier bulb, on which the railway is built, need not necessarily be broken up. The railway engineers could not have avoided this stretch of track upon the glacier, without prohibitive expense of rock cuts, a tunnel, and two expensive bridges across Copper River. The most serious error in the railway building was the stripping off of all vegetation from the right of way. If the alders had been left growing close up to the railway the slumping of the ice beneath and close to the track would have been much less. Even now it would probably pay to plant the right of way with new shrubs so that further slumping beside the track may be decreased. It seems likely that renewal of ballast beneath the rails will eventually bury the ice there so deeply that slumping will cease. i Since making this prediction of imminent advance of Allen Glacier the junior author has been informed by Mr. Caleb Corser of the Copper River and Northwestern Railway that during the summer of 1912 the glacier commenced to move forward. The southern border was crevasaed and broken in September, 1912, and showed dear blue ice where there was smooth, moraine-veneered ice in 1911. In the northern, moraine-covered margin of the bulb there was much thickening and crevassmg and an advance which is said to amount to } mile. The largest glacial stream on the northern side of Allen Glacier left the channel which it occupied from 1909 to 1911 and flowed in a new channel, in 1912, about a mile farther west. Its volume was greatly increased and the stream shifted frequently, causing much trouble in the portion of the raflway grade on the northern alluvial fan of outwash gravels.